HCA is an important technique for repair of complex congenital heart lesions and operations on the aortic arch and thoracic aorta [1
]. It provides a bloodless operative field, unobstructed by vascular clamps and cannulae. However, the central nervous system is exquisitely sensitive to ischemia, limiting the duration of safe arrest. Neurological injury occurs when the duration of arrest exceeds 45–60 minutes [4
]. Clinical sequelae include seizures, choreoathetosis, learning and memory deficits, and impaired intellectual development [4
]. Clinically apparent neurologic damage occurs in up to 12% of children and 15% of adults [4
Neuronal cell death causes HCA induced neurologic injury. Necrosis is one neuropathologic hallmark of HCA induced injury; however, we have previously shown that neuronal apoptosis is an additional cause of neurologic injury after HCA [9
]. We demonstrated that neuronal apoptosis is a process occurring early, peaking 8 hours, diminishing by 20 hours, and disappearing by 72 hours after HCA. This time frame may be different in other animal models, such as neonatal pigs, and the acute nature of this time frame may also be influenced by the degree and severity of ischemia, where shorter durations of HCA may show a longer time till onset of apoptosis [15
Previous work in our laboratory has demonstrated the utility of glutamate receptor antagonists in ameliorating necrosis associated with HCA at 72 hours [16
]. In contrast, one study showed no neuroprotective effect of memantine, an NMDA receptor antagonist in a porcine model of 75 minutes of HCA at 20°C after 7 days [18
]. They found no significant changes in histology or behavioral scoring; however, since they did not investigate changes in nitric oxide, it is uncertain whether the dose response curve of memantine was sufficient to reduce nitric oxide production. In this study, we investigated the mechanism of apoptosis which occurs more acutely in our HCA model using an NMDA receptor antagonist, MK-801. We used the same dose of MK-801 as Dr. Redmond’s chronic study examining necrosis [10
], and demonstrated not only inhibition of apoptosis after HCA but also near complete suppression of nitric oxide production.
In our prior microdialysis study, we examined changes in extracellular amino acid concentrations in the brain associated with neuronal apoptosis after hypothermic circulatory arrest [19
]. Glutamate levels increased significantly during arrest, reperfusion CPB, and recovery, supporting the hypothesis of glutamate excitotoxicity. In addition, citrulline levels, reflective of NO production, increased significantly during reperfusion CPB and recovery [19
]. In this study, MK-801, a noncompetitive antagonist, did not change glutamate levels; however, it did significantly reduce NO production, supporting the upstream actions of glutamate. MK-801 also reduced glycine levels, a requisite co-agonist of glutamate on the NMDA receptor.
We have demonstrated that glutamate excitotoxicity mediates neuronal apoptosis in this model of HCA. Apoptosis was pathologically defined by cellular shrinkage and nuclear chromatin condensation and aggregation on H&E and electron microscopy. Apoptosis was the morphologic representation of regularly cut nucleosome length DNA, and the TUNEL assay of nick-end labeling demonstrated cells with DNA fragmentation. Subsequent work in Dr. Johnston’s laboratory has demonstrated good correlation of TUNEL positive staining with activated caspase staining in the dentate gyrus for apoptosis. Increased glutamate via its effects on NO production may result in DNA strand breaks by base deamination to result in apoptosis.
Glutamate Excitotoxicity and Cell Death Mechanism
Neuronal injury may be caused by overstimulation of excitatory amino acid receptors, including glutamate and aspartate [20
]. This excitotoxicity is predominantly mediated by calcium influx through ionic channels of activated glutamate receptors. Hypoxia and ischemia result in overaccumulation of the excitatory amino acid, glutamate. Glutamate, the principal neurotransmitter of the brain, is responsible for many physiologic functions, including cognition, memory, movement, and sensation. Pathophysiologically, excessive glutamate activates N
-methyl-D-aspartate (NMDA), α-amino-3-hydroxy-5-methyl-4-isoxazolepropionate (AMPA), and kainate glutamate receptors. Glutamate along with co-agonist glycine stimulates NMDA receptors to increase intracellular calcium, which triggers a cascade of intracellular reactions, activating phospholipases, proteases, protein kinases, phosphatases, and nitric oxide synthase (NOS). NOS causes increased nitric oxide production, which may damage DNA by base deamination to result in DNA strand breaks. Damaged DNA activates poly(adenosine 5′-diphosphoribose) polymerase (PARP) to add dATP to the ends of nicked DNA, resulting in depletion of energy sources from the cell. These processes ultimately lead to cell death, which can be necrotic or apoptotic in nature [22
Apoptosis, or programmed cell death, is a tightly regulated process, requiring energy, macromolecular synthesis, and gene transcription [21
]. It results in nonrandom oligonucleosomal length DNA fragmentation. A number of genes have been implicated in the programmed cell death process, including antiapoptotic genes, i.e. bcl-2, and proapoptotic genes, i.e. bax and interleukin converting enzyme (ICE) family of cysteine proteases [21
]. However, what shifts the balance of antiapoptosis to proapoptosis is unknown. Furthermore, why cells die via apoptosis vs. necrosis is unknown. Evidence suggests that any insult below the threshold to cause necrosis may result in apoptosis, so that mild and intense insults result in apoptosis and necrosis, respectively [21
]. Initiation of the apoptotic pathways appears to involve increasing intracellular calcium, acidosis, reactive oxygen species, and Fas receptor activation [21
]. Reduction in neuronal apoptosis has primarily focused on calpain and caspase inhibition and antioxidants [26
Neuronal Cell Death after HCA
In hypothermic circulatory arrest, glutamate excitotoxicity mediates both apoptosis and necrosis [27
]. How glutamate triggers one cell to undergo apoptosis, while another necrosis is not known. Severe ischemia with large increases in glutamate may result in necrosis, characterized by cellular swelling and lysis, while milder insults cause apoptosis. Greater insults may generate more superoxide anion to convert nitric oxide to peroxynitrite, which leads to lipid peroxidation and cellular lysis. Milder insults may generate nitric oxide, but less peroxynitrite, to alter the oxidative state of the cell and shift the balance from antiapoptotic to proapoptotic forces.
In this study, we have demonstrated that NMDA receptor antagonism with MK-801 significantly reduced downstream nitric oxide production with no effect on glutamate concentrations. This reduction in nitric oxide resulted in significant reduction in neuronal apoptosis after HCA. Apoptotic cell death is significant not only for its role in neurologic injury after HCA, but also for its role in hypoxia, ischemia, Huntington’s, Alzheimer’s, and physiological neural development. Strategies for cerebral protection after HCA require amelioration of both forms of neuronal cell death, necrosis and apoptosis. Glutamate receptor antagonists in HCA are efficacious in reducing both necrosis and apoptosis; however, long-term efficacy and outcomes are uncertain due to the complicating factors of their own neurologic side effects, including problems with memory and cognition. Clinically safe glutamate receptor antagonists with minimal side effects are challenging to develop, but would be beneficial in reducing neurologic complications after HCA.